The use of Artificial Intelligence (AI) and the programming language Ada to help a satellite recover from selected failures that could lead to mission failure are described. An unmanned satellite will have a separate AI subsystem running in parallel with the normal satellite subsystems. A satellite monitoring subsystem (SMS), under the control of a blackboard system, will continuously monitor selected satellite subsystems to become alert to any actual or potential problems. In the case of loss of communications with the earth or the home base, the satellite will go into a survival mode to reestablish communications with the earth. The use of an AI subsystem in this manner would have avoided the tragic loss of the two recent Soviet probes that were sent to investigate the planet Mars and its moons. The blackboard system works in conjunction with an SMS and a reconfiguration control subsystem (RCS). It can be shown to be an effective way for one central control subsystem to monitor and coordinate the activities and loads of many interacting subsystems that may or may not contain redundant and/or fault-tolerant elements. The blackboard system will be coded in Ada using tools such as the ABLE development system and the Ada Production system.